From 4458978d3871909c047319aba1102f32e6b51349 Mon Sep 17 00:00:00 2001
From: Eric Mill <eric.mill@gsa.gov>
Date: Fri, 13 Jul 2018 12:55:07 -0400
Subject: [PATCH] start measuring uswds presence

---
 scanners/uswds.js | 15 ++++++++++++++-
 scanners/uswds.py |  4 ++--
 2 files changed, 16 insertions(+), 3 deletions(-)

diff --git a/scanners/uswds.js b/scanners/uswds.js
index 45d52fde..1703eead 100644
--- a/scanners/uswds.js
+++ b/scanners/uswds.js
@@ -36,14 +36,27 @@ module.exports = {
       else throw exc;
     }
 
+    // Do a quick text match on the HTML to find out of date USWDS
+    // banner text.
     const html = await page.content();
-
     data.banner_bad_text = hasBadText(html);
 
+    // Search DOM for evidence of USWDS being present at all.
+    data.present = await uswdsPresent(page);
+
     return data;
   }
 }
 
+// Simple text match on some old text.
 var hasBadText = (html) => {
   return (html.search("Federal government websites always use a .gov or .mil domain.") >= 0)
 };
+
+
+// Checks for any element with a class with "usa-" in it.
+// TODO: Should "usa-" at least be at the start of the string?
+var uswdsPresent = async (page) => {
+  var match = await page.$('[class*=usa-]');
+  return (match != null);
+};
diff --git a/scanners/uswds.py b/scanners/uswds.py
index ee848175..f553558b 100644
--- a/scanners/uswds.py
+++ b/scanners/uswds.py
@@ -51,13 +51,13 @@ def to_rows(data):
 
     return [[
         data['url'],
-        data.get('banner_present'),
+        data.get('present'),
         data.get('banner_bad_text'),
     ]]
 
 
 headers = [
     'Scanned URL',
-    'USWDS Banner Present',
+    'USWDS Present',
     'USWDS Bad Banner Text'
 ]